International Journal of Computer Networks (IJCN) Volume 2, Issue 4, 2010

8/8/2019 International Journal of Computer Networks (IJCN) Volume 2, Issue 4, 2010

1/30


2/30

International Journal of

Computer Networks (IJCN)

Volume 2, Issue 4, 2010

Edited ByComputer Science Journals

www.cscjournals.org


3/30

Editor in Chief Dr. Min Song

International Journal of Computer Network

(IJCN)

Book: 2010 Volume 2, Issue 4

Publishing Date: 30-10-2010

Proceedings

ISSN (Online): 1985-4129

This work is subjected to copyright. All rights are reserved whether the whole or

part of the material is concerned, specifically the rights of translation, reprinting,

re-use of illusions, recitation, broadcasting, reproduction on microfilms or in any

other way, and storage in data banks. Duplication of this publication of parts

thereof is permitted only under the provision of the copyright law 1965, in its

current version, and permission of use must always be obtained from CSC

Publishers. Violations are liable to prosecution under the copyright law.

IJCN Journal is a part of CSC Publishers

http://www.cscjournals.org

IJCN Journal

Published in Malaysia

Typesetting: Camera-ready by author, data conversation by CSC Publishing

Services CSC Journals, Malaysia

CSC Publishers


4/30

Editorial Preface

The International Journal of Computer Networks (IJCN) is an effectivemedium to interchange high quality theoretical and applied research in thefield of computer networks from theoretical research to application

development. This is the fourth issue of volume second of IJCN. The Journalis published bi-monthly, with papers being peer reviewed to highinternational standards. IJCN emphasizes on efficient and effective imagetechnologies, and provides a central for a deeper understanding in thediscipline by encouraging the quantitative comparison and performanceevaluation of the emerging components of computer networks. Some of theimportant topics are ad-hoc wireless networks, congestion and flow control,cooperative networks, delay tolerant networks, mobile satellite networks,multicast and broadcast networks, multimedia networks, networkarchitectures and protocols etc.

IJCN give an opportunity to scientists, researchers, engineers and vendors toshare the ideas, identify problems, investigate relevant issues, sharecommon interests, explore new approaches, and initiate possiblecollaborative research and system development. This journal is helpful forthe researchers and R&D engineers, scientists all those persons who areinvolve in computer networks in any shape.

Highly professional scholars give their efforts, valuable time, expertise andmotivation to IJCN as Editorial board members. All submissions are evaluatedby the International Editorial Board. The International Editorial Board ensuresthat significant developments in computer networks from around the worldare reflected in the IJCN publications.

IJCN editors understand that how much it is important for authors andresearchers to have their work published with a minimum delay aftersubmission of their papers. They also strongly believe that the directcommunication between the editors and authors are important for thewelfare, quality and wellbeing of the journal and its readers. Therefore, allactivities from paper submission to paper publication are controlled throughelectronic systems that include electronic submission, editorial panel andreview system that ensures rapid decision with least delays in the publicationprocesses.

To build its international reputation, we are disseminating the publicationinformation through Google Books, Google Scholar, Directory of Open AccessJournals (DOAJ), Open J Gate, ScientificCommons, Docstoc and many more.Our International Editors are working on establishing ISI listing and a goodimpact factor for IJCN. We would like to remind you that the success of our

journal depends directly on the number of quality articles submitted forreview. Accordingly, we would like to request your participation by


5/30

submitting quality manuscripts for review and encouraging your colleagues tosubmit quality manuscripts for review. One of the great benefits we canprovide to our prospective authors is the mentoring nature of our reviewprocess. IJCN provides authors with high quality, helpful reviews that areshaped to assist authors in improving their manuscripts.

Editorial Board MembersInternational Journal of Computer Networks (IJCN)


6/30

Editorial Board

Editor-in-Chief (EiC)

Dr. Min Song

Old Dominion University (United States of America)

Associate Editors (AEiCs)

Dr. Qun Li

The College of William and Mary (United States of America)

Dr. Sachin Shetty

Tennessee State University (United States of America)

Dr. Liran Ma

Michigan Technological University (United States of America)[

Dr. Benyuan LiuUniversity of Massachusetts Lowell (United States of America)

Editorial Board Members (EBMs)

Dr. Wei Chen

Tennessee State University (United States of America)Dr. Yu CaiMichigan Technological University (United States of America)

Dr. Ravi Prakash RamachandranRowan University (United States of America)Dr. Bin WuUniversity of Waterloo (Canada)

Dr. Jian RenMichigan State University (United States of America)Dr. Guangming Song

Southeast University (China)

Dr. Tian-Xiao HeIllinois Wesleyan University (United States of America)Dr. Jiang Li

Howard University (China)Dr. Baek-Young ChoiUniversity of Missouri Kansas City (United States of America)

Dr. Fang Liu

University of Texas at Pan American (United States of America)Dr. Lawrence MillerThe University of Toledo (United States of America)

Dr. Enyue LuSalisbury University (United States of America)Dr. Chunsheng Xin

Norfolk State University (United States)

Associate Professor. Wenbin JiangHuazhong University of Science and Technology (China)


7/30

Associate Professor. Xiuzhen Cheng

The George Washington University (United States)

Dr. Imad JawharUnited Arab Emirates University (United Arab Emirates)Associate Professor. Lawrence Yeung

The University of Hong Kong (Hong Kong)

Dr. Yong CuiTsinghua University (China)

Dr. Wei ChengThe George Washington University (United States of America)

Dr. Filip CuckovOld Dominion University (United States of America)

Dr. Zhong Zhou

University of Connecticut (United States of America)Dr. Mukaddim Pathan

CSIRO-Commonwealth Scientific and Industrial Research Organization (Australia)Associate Professor. Cunqing Hua

Zhejiang University (China)


8/30

International Journal of Computer Network (IJCN) Volume (2): Issue (4)

Table of Content

Volume 2, Issue 4, October 2010

Pages

173 - 180

181 - 189

The Design of a Simulation for the Modeling and Analysis of Public

Transportation Systems as Opportunistic Networks

Rotimi Iziduh, Jamahrae Jackson, Howard Sueing, A. Nicki

Washington, Robert Rwebangira, Legand Burge

Queue Size Trade Off with Modulation in 802.15.4 for Wireless

Sensor Networks

Sukhvinder S Bamber, Ajay K Sharma


9/30

Iziduh, Jackson, Sueing, Washington, Rwebangira & Burge

International Journal of Computer Networks (IJCN), Volume(2): Issue (4) 173

The Design of a Simulation for the Modeling and Analysis of PublicTransportation Systems as Opportunistic Networks

Rotimi Iziduh [email protected] of Systems and Computer ScienceHoward University Washington,DC 20059 USA

Jamahrae Jackson [email protected] of Systems and Computer ScienceHoward University Washington,DC 20059 USA

Howard Sueing [email protected] of Systems and Computer ScienceHoward University Washington,

DC 20059 USA

A. Nicki Washington [email protected] of Systems and Computer ScienceHoward University Washington,DC 20059 USA

Robert Rwebangira [email protected] of Systems and Computer ScienceHoward University Washington,DC 20059 USA

Legand Burge [email protected] of Systems and Computer ScienceHoward University Washington,DC 20059 USA

Abstract

Vehicular ad-hoc networks, when combined with wireless sensor networks, are used ina variety of solutions for commercial, urban, and metropolitan areas, includingemergency response, traffic, and environmental monitoring. In this work, we modelbuses in the Washington, DC Metropolitan Area Transit Authority (WMATA) as a

network of vehicular nodes equipped with wireless sensors. A simulation tool wasdeveloped, to determine performance metrics such as end-to-end packet delivery delay.

Keywords: Opportunistic networks, Vehicular networks, Simulation, Network simulation


10/30



1. INTRODUCTIONMobile ad hoc networks (MANET) have provided technological connectivity in areas where variousconstraints, including environmental, financial, cultural, time, and government prohibited theestablishment of infrastructure-based networks. Nodes may be static or mobile, leading to a dynamicnetwork topology. Routing of data occurs as nodes relay information to each other. Traditional ad hocrouting protocols assume the network is fully connected. In addition, the end-to-end source-destination

path is assumed to be known prior to transmission.

The need for increased connectivity extends from urbanized areas to remote and rural areas previouslyunreachable via standard telecommunication networks. In either of these cases, the establishment or useof an infrastructure-based network is not always feasible, due to various constraints, including time,financial, cultural, government, and environmental. In addition, certain catastrophic events can renderinfrastructure networks useless.

Opportunistic or disruption tolerant networks (DTN) are special types of MANETs where no end-to-endpath exists between source and destination nodes, due to a number of potential factors, including nodemobility, physical obstructions, etc. Packet transmission occurs in a store-and-forward fashion, wherenodes relay packets to neighboring nodes as they come in contact with each other, until the packetultimately reaches its destination. As a result, packets must endure longer delays.

Vehicular ad-hoc networks (VANETs) are a special type MANET where cars or buses are equipped withdevices that allow them to communicate with each other and any stationary equipment they may pass.These vehicles, referred to as nodes, are restricted to movement on streets or designated paths. In amajor metropolitan area, public transportation systems can be utilized to provide opportunistic routing anddelivery of data via buses. When equipped with wireless sensors, these networks can be used for anumber of purposes, including health, environmental, habitat, and traffic monitoring, emergencyresponse, and disaster relief [4, 10, 11, 15, 16, 20, 22].

In this work, we develop a simulation tool for modeling and analyzing a DTN comprised of buses in apublic transportation system. Using real bus information and schedules, the simulation provides a realisticmodel of the entire network. This tool can be used to study various routing protocols and networkperformance metrics, such as end-to-end packet delay, packet copy distribution, and more. In addition,we provide a web-based front-end, using the Google Maps API, that provides a user-friendly interface for

updating the network to account for a number of parameters and conditions, including inclement weather,traffic congestion, and other adverse conditions.

We note that, while this work uses the Washington Metropolitan Area Transit Authority (WMATA) system,the simulation can model any public transportation system that subscribes to the defined specification. Itis the ultimate goal that this simulation will be used not only to study the use of the public transportationsystems of cities for various societal and research purposes, but also to provide a means for anyorganization or individual to utilize this tool to gather relevant data.

The remainder of this work is organized as follows. In section 2, we discuss related work on DTNsimulation models. In section 3, we present the network model. In section 4, we present the simulationmodel and web-based front end. In section 5, we present numerical results and a snapshot of oursimulation application. In section 6, we conclude our findings.

2. RELATED WORKOpportunistic or delay tolerant networks (DTN) have been suggested as a viable solution for a number ofnon-traditional mobile ad-hoc networks. These include providing connectivity in rural or remote areas,wildlife tracking and monitoring, and military battlefields, to name a few. A large majority of the work hasfocused on the development and analysis of routing protocols to measure a number of performancemetrics, including end-to-end delay and packet copy distribution.


11/30



Recently, work in the area of DTN has shifted to include urban environments and the capabilities in theseareas [2, 8, 10, 15, 19, 23]. Specifically, the use of vehicular nodes has been studied. These networks areassumed to perform a number of tasks, including traffic and environment monitoring, and emergencyresponse, and disaster recovery. A large amount of work has focused on various routing strategies forthese networks [1, 2, 3, 5, 7, 9, 11, 15, 16, 17, 18].

In these networks, there are a number of attempts to model various protocols using testbeds [12, 13, 14,15, 23]. These testbeds range from small networks of robots emulating the movement of vehicles toactual buses equipped with processing capabilities. In each of these cases, there are limitations to theimplementation. Mainly, the size of the actual networks can make the creation of an exact replicaextremely difficult, if not impossible. Second, the implementation of these studies on the actual networkcan be difficult to implement, due to financial, regulatory, and time constraints.

Simulations are a viable solution to modeling and analyzing opportunistic networks exploiting vehicularnodes in urban areas [15, 16, 19]. These networks can be analyzed, in full, prior to implementation. Thebenefit in this research is the ability to study the performance of the network using a number of protocolsand parameters. While these are easier to implement than physical testbeds, creating such a large-scalesimulation, that incorporates a large number of nodes and data movement can quickly become complex.This requires adequate emulation of vehicular movement, data transmission, schedules, and more.

To the best of our knowledge, there is still much open research in the area of simulations of DTN in urbanenvironments. Specifically, the use of public transportation (i.e. buses, subway, etc.) as vehicular nodeshas only recently begun to receive attention. The complexity of such networks, due to the aforementionedschedules, number of nodes, etc. can make this type of network difficult to accurately simulate.

In addition, the aforementioned simulations are designed to address a specific network representation.Our simulation allows for various public transportation networks to be studied using the Google TransitFeed Specification. Simulation parameters do not require modification when studying various cities.

Our research focuses on developing a simulation tool that can be used to study DTN in urbanenvironments exploiting the public bus system of any city. The novelty of our research is the developmentof a tool that can be used to study various protocols and metrics of interest by providing a common formatfor modeling any city, using a Google-developed specification for transit data. While simulations have

been developed, to the best of our knowledge, our work is the first that incorporates real data, includingschedules, sing Google Transit Feed Specification. The benefit of using this that the tool can be used tostudy any city whose public transportation information is provided via the specification. Furthermore, theweb-based front-end provides a simplified mechanism for manipulating and executing the simulation.

3. NETWORK MODELThe network is composed of all streets that comprise the WMATA public transportation grid, includingWashington, DC-proper and adjacent cities in both Maryland and Virginia. A node in the network isrepresented by a bus. Each station and node is assumed to be equipped with processing capabilities anda small buffer. Each bus belongs to a bus (node) line, which has a pre-determined path comprised of aset of streets. We note that a single line contains multiple buses traveling in opposite directions, referredto as upstream and downstream. In addition, every bus on every line has an expected arrival/departure

time to/from each designated stop along the line.

A stop is defined as a stationary bus stop or base station, where data collection/dissemination activitiestake place. We assume each stop is equipped with the necessary equipment (e.g. sensors, etc.) to collectand store data. At any stop, a packet is randomly generated that is destined for another stop. The packetis transmitted to the first node that reaches the stop after this generation. As the carrier node travelsthroughout the network, it transmits the packet to any node it encounters that is within transmissionrange. A packet is delivered once it reaches the destination stop.In this work, we make the following assumptions:


12/30



Packets are originated at and delivered to stops in the network. Buses (nodes) serve asintermediate carriers for relaying information from a source (stop) to destination (stop).

A packet can be transmitted to a node or line that previously carried the packet.

If two nodes are within communication range of the current carrier node, then the new carrier israndomly selected.

At the end of the last bus route, buses store all packets in their respective buffers until the nextdays routes begin.

Packets arriving to nodes with full buffers are undeliverable. However, the transmitting noderetains copies of the packets.

4. SIMULATION MODELTo test our simulation, we used actual bus information from the Washington Metropolitan Area TransitAuthority (WMATA). This information includes a total of approximately 1,400 buses on 350 different buslines over approximately 80 sq. miles. We note that each bus line contains more than one bus in eachdirection.

In order to accurately and efficiently obtain and analyze real WMATA information, we used the General

Transit Feed Specification (GTFS). GTFS is a Google-defined specification that provides a commonformat for mapping a citys public transportation with its associated geographic info using the GoogleMaps API [6, 21]. Using GTFS, the public transportation agency of any city can prepare a data feedaccording to specifications, validate it, and enroll in a partnership with the Google Transit team to launchthe feed in GTFS. Submitted feed information includes subway, bus, and train info. The associated datais then displayed in Google Maps. Information provided via GTFS includes parameters such asstation/stop name, longitude, latitude, bus/subway stops and lines, etc. A number of public transportationagencies across the country and globe currently have feeds in GTFS, including major metropolitan areassuch as San Francisco, Boston, Philadelphia, Washington, DC, and New York.

Using a custom, Java-based discrete-event simulator, we model and simulate the movement of busesand data in the network. We use GTFS, JavaScript, XML, and the Google Maps API to build a custom,web-based front-end for our simulation. This front-end was developed to provide an alternative view of

the simulation results. Users can not only view the resulting path that a delivered packet traverses, butalso manipulate specific input parameters, such as number of network packets, start time, source-destination pairs, and adverse conditions (inclement weather, accidents, etc.) within a user-friendlyenvironment. The front-end was developed with the goal of providing various agencies and authoritiesstudying the use of public transportation for various research or application purposes could use thesimulation, combined with the corresponding GTFS feed for a specific city, and easily manipulate thesimulation, regardless of their level of knowledge regarding the simulation design and implementation.l

Using the information provided by GTFS and our discrete-event simulator, packets are randomlygenerated at stops, and transferred by nodes to destination stops. Each node has a small buffer thatallows for the storing of packets in transit. The exchange of a data packet occurs when two nodes arewithin 500 ft. of each other.

5. NUMERICAL RESULTSIn this work, we simulate a seven-day time period in the WMATA. Packets are randomly generatedthroughout the network at various sources and at various times of day. In addition, we assume thatpackets are routed according to flooding. At the end of each day, all buses with packets in their respectivebuffers store these packets until the next mornings route begins. For our work, we use a maximum nodebuffer size of 12 packets and a maximum number of 10 unique packets destined for different source-destination packets in the network. We note that, while 10 unique packets are generated, there aremultiple copies of these 10 packets present throughout the network.


13/30



Figure 1 presents a snapshot of the current web-based front-end. It should be noted that, in addition tobeing able to map the resulting path traversed, the simulation also allows users to input a number ofparameters to manipulate the simulation, including weather conditions, rush hour starting time, packetsize, etc.

FIGURE 1: Web-Based Front End

Figure 2 presents the average end-to-end delivery delay of packets as a function of the size of the buffer.We note that, as expected, the delay is reduced when the buffer is introduced. In our previous preliminarywork, we noted that a network with no store-and-forward capabilities (i.e. buffer size of 0) was significantlyhigher [19]. For this reason, we only focus on buffer capabilities in this work.

Delay vs. Buffer Size

0

0.5

1

1.5

2

2.5

0 2 4 6 8 10 12

Buffer Size (# Packets)

Delay(hrs)

FIGURE 2: Packet Delay as a Function of Buffer Size


14/30



We also note that the delay varies slightly between 2 to 10 packets buffer size. We attribute this to asmaller number of network packets in the network. With increased number of unique packets and copiesin the network, the delay variation is expected to be greater.

Figure 3 presents the average end-to-end delay as a function of the number of unique packets in thenetwork. We note an interesting trend in this figure. As expected, as the number of packet copiesincreases, the end-to-end delay increases. We note the dramatic reduction in delay from 1 to 2 packets.This is due to the fact that multiple copies of an individual packet are now in the network, which allows forthe probability of quicker delivery.

Delay vs. Number of Packets

0

5

10

15

20

25

0 2 4 6 8 10 12

Number of Packets

Delay(hours)

FIGURE 3: Packet Delay as a Function of Number Packets

6. CONCLUSIONIn this work, we developed a discrete-event simulation to represent an opportunistic network comprised ofvehicular nodes. A web-based front end was also developed via the Google Maps API, to allow a user-friendly method of manipulating network and simulation parameters. The simulation used real businformation from the Washington Metropolitan Area Transit Authority (WMATA). However, this simulationcan easily analyze any system utilizing the General Transit Feed Specification (GTFS).

Currently, we are extending the capabilities of the simulation to include more unique source-destinationpairs, complete parameter specification on the web-based front end, the incorporation of mobile basestations across the city, and the inclusion of additional node traffic (i.e. subway, people, cars, etc.).

We are also working on developing a model for approximating the analysis of the network. This work canultimately be used to assist metro authorities in various cities with addressing optimization problems, suchas costs, routing issues, and resource allocation. It can also be used to model the performance of varioustypes of algorithms and protocols on such networks, including those used for emergency response,disaster relief, environmental monitoring, and more.


15/30



7. REFERENCES[1] Conan, V. et. al.Fixed Point Opportunistic Routing in Delay Tolerant Networks. Journal on SelectedAreas in Communication: Special Issue on Delay Tolerant Networks, 26(5):773-782, 2008

[2 ] Dias, J. et. al. Creation of a Vehicular Delay-Tolerant Network Prototype. In Proceedings ofEngenharia - Jornadas da Universidade da Beira Interior, Covilha, Portugal, 2009

[3] Doria, A., Uden, M., and Pandey, D. Providing Connectivity to the Saami Nomadic Community. InProceedings of the 2

ndInternational Conference on Open Collaborative Design for Sustainable

Innovation, Bangalore, India, 2002

[4] Echelon Corporation Buses in Italy Stop for LonWorks Networks [Online]. Available at:http://www.echelon.com. [Accessed from the Echelon Networks website]

[5] Eisenman, S, et. al. Techniques for Improving Opportunistic Sensor Networking Performance.Lecture Notes In Computer Science,5067:157 175, 2008

[6] Google. General Transit Feed Specification [Online]. Available at:http://code.google.com/transit/spec/transit_feed_specification.html [Accessed from the Google website]

[7] Groenevelt, R., Nain, P., Koole, G. The Message Delay in Mobile Ad Hoc Networks. Performance62(1-4):210-228, 2005

[8] Lane, N., et. al. Urban Sensing: Opportunistic or Participatory?In Proceedings of the 9th

Workshopon Mobile Computing Systems and Applications, Napa Valley, CA, USA, 2008

[9] Leguay, J. et. al. Evaluating MobySpace Based Routing Strategies in Delay Tolerant Networks.Wireless Communication and Mobile Computing, 7(10):1171-1182, 2007

[10] Kohl, G. Going Mobile with Video Surveillance on Chicagos Bus System [Online]. Available at:http://www.securityinfowatch.com/root+level/1284404 [Accessed from the Security Info Watch website]

[11] Li, X., et. al. DTN Routing in Vehicular Sensor Networks. In Proceedings of IEEE Global

Telecommunications Conference, New Orleans, LO, USA 2008

[12] Murty, Rohan., et. al. CitySense: An Urban-Scale Wireless Sensor Network and Testbed. InProceedings of IEEE Conference on Technologies for Homeland Security, Waltham, MA, USA, 2008

[13] Ormont, Justin., et. al. A City-Wide Veheicular Infrastructure for Wide-area WirelessExperimentation. In Proceedings of International Conference on Mobile Computing and Networking, SanFrancisco, CA, USA, 2008

[14] Ormont, Justin., et. al. Continuous monitoring of wide-area wireless networks: data collection andvisualization. In Proceedings of ACM SIGMETRICS Performance Evaluation Review, 2008

[15] Sede, Michel., et. al. Routing in Large-Scale Buses Ad Hoc Networks. In Proceedings of IEEE

Wireless Networking Conference, Las Vegas, NV, USA, 2008

[16] Seet, Boon-Chong., et. al. A-STAR: A Mobile Ad Hoc Routing Strategy for Metropolis VehicularCommunications. In Proceedings of Networking, 2004

[17] Seth, A., et. al. Low-Cost Communication for Rural Internet Kiosks Using Mechanical Backhauls. InProceedings of ACM MobiCom Los Angeles, CA, USA, 2006[18] Shah., R., et. al. Data MULEs: Modeling a Three-tier Architecture for Sparse Sensor Networks. InProceedings of IEEE SNPA Workshop, 2003


16/30



[19] Sueing, H. et. al. The Modeling and Analysis of the Washington Metropolitan Area Bus Network. InProceedings of the 2010 Modeling, Simulation, and Visualization Conference, Las Vegas, NV, USA, 2010

[20] Vahdat, A., and Becker, D., et. al. Epidemic Routing for Partially-Connected Ad Hoc Networks.Tech. Rep. CS-2000006, 2000

[21] Washington Metropolitan Area Transit Authority. Developer Resources [Online]. Available at:http://www.wmata.com/rider_tools/developer_resources.cfm [Accessed from the Washington MetropolitanArea Transit Authority website]

[22] Zhang, Z., et. al. Routing in Intermittently Connected Mobility Ad Hoc Networks And Delay TolerantNetworks: Overview and Challenges. IEEE Communication Surveys and Tutorials, 8(1):24-37, 2006

[23] Zhang, X., et al. Study of Bus-based Disruption-Tolerant Network: Mobility Modeling and Impact onRouting. In Proceedings of ACM MobiCom, New York, NY, 2007


17/30

Sukhvinder S Bamber and Ajay K Sharma


Queue Size Trade Off with Modulation in 802.15.4 for WirelessSensor Networks

Sukhvinder S Bamber [email protected] of Computer Science & EngineeringNational Institute of Technology,Jalandhar, 144011, India

Ajay K Sharma [email protected] of Computer Science & EngineeringNational Institute of Technology,Jalandhar, 144011, India

Abstract

In this paper we analyze the performance of 802.15.4 Wireless Sensor Network

(WSN) and derive the queue size trade off for different modulation schemes like:Minimum Shift Keying (MSK), Quadrature Amplitude Modulation of 64 bits(QAM_64) and Binary Phase Shift Keying (BPSK) at the radio transmitter ofdifferent types of devices in IEEE 802.15.4 for WSN. It is concluded that if queuesize at the PAN coordinator of 802.15.4 wireless sensor network is to be takeninto consideration then QAM_64 is recommended. Also it has been concludedthat if the queue size at the GTS or Non GTS end device is to be considered thenBPSK should be preferred. Our results can be used for planning and deployingIEEE 802.15.4 based wireless sensor networks with specific performancedemands. Overall it has been revealed that there is trade off for using variousmodulation schemes in WSN devices.Keywords:WSN, Queue Size, BPSK, MSK, QAM_64.

1. INTRODUCTION

The IEEE 802.15.4 protocol is an industrial standard for Low-Rate Wireless Personal AreaNetwork (LR-WPAN) architectures. As the primary application domain wireless sensor networkapplications in industrial environments can be identified. LR-WPAN is intended to become anenabling technology for WSNs. In contrast to Wireless Local Area Networks (WLAN), which isstandardized by IEEE 802.11 family, LR-WPAN stresses short-range operation, low-data-rate,

energy-efficiency and low-cost. An example is Zigbee, which is an open specification built on theLR-WPAN standard and focuses on the establishment and maintenance of LR-WPANs forwireless sensor networks.

The choice of the digital modulation scheme significantly affects the characteristics, performanceand resulting physical realization of wireless sensor communication system derived from802.15.4. There is no universal best choice of the modulation scheme, but depending on thephysical characteristics of the channel, parametric optimizations and required level ofperformance some will prove better fit than the others. The 802.15.4 is an IEEE standard,


18/30



targeting a set of applications that require simple wireless connectivity, high throughput, very lowpower consumption and lower module cost. Its objective is to provide low complexity, cost andpower for wireless sensor connectivity among inexpensive, fixed, portable and moving devices.

A lot of work on 802.15.4 has been reported by the various researchers [1-22]. The performanceissues like: Delay; Throughput evaluation of GTS mechanism have been reported in [1].Researchers have also studied adaptive algorithm for mapping channel quality information tomodulation and coding schemes [3]. Researchers have also tried to study performance tradeoffwith adaptive frame length and modulation in wireless network [4]. Some researchers havestudied suboptimum receivers for DS-CDMA with BPSK modulation [5]. Researchers have alsoinvestigated voice and data transmission technique using adaptive modulation [6]. Manyresearchers have studied how to use queues to improve the performance of TCP [10]. Somehave studied queues o dynamically allocate the channels for real-time and non-real-time traffic incellular networks [12]. Few have studied queues for energy and QoS tradeoff for contention-based wireless sensor networks [15]. Some have worked on how to stabilize queues in large-scale networks [16]. Few researchers have studied the queues for controlling the power inwireless communication networks [20]. But none of the researchers have reported theperformance comparison using different modulation schemes for 802.15.4 based on queue size.This paper proposes the comparison of different modulation schemes (QAM_64, MSK, BPSK)based on queue size to determine the suitability of 802.15.4 network.

Section [1] gives the brief introduction. Section [2] constitutes the system description whichcontains node model, process model, and parametric tables of the model. Section [3] shows theresults and discussions derived from the experiments carried out on 802.15.4 for differentmodulation schemes. Finally Section [4] concludes the paper.

2. SYSTEM DESCRIPTION

The simulation model implements physical and medium access layers defined in IEEE 802.15.4standard. The OPNET

Modeler 14.5 is used for developing 802.15.4 wireless sensor network.

(a) (b)


19/30



(c)Figure 1:Network Scenarios (a) BPSK (b) MSK (c) Quadrature (QAM_64)

Figure 1 shows three different Scenarios: BPSK, MSK and QAM_64. BPSK Scenario as shown inFigure 1(a) contains one PAN Coordinator, one analyzer and thirty two end devices out of whichsixteen are Guaranteed Time Slots (GTS) enabled and rest are non GTS devices. PANCoordinator is a fully functional device which manages whole functioning of the network. Analyzeris a routing device which routes the data between PAN coordinator and the End Devices. EndDevices are the fixed stations that communicate with the PAN Coordinator in Peer to Peer mode,support GTS and non GTS traffic respectively. Similar Scenarios have been created for MSK andQAM_64 as shown in figure 1 (b & c).

Figure 2 shows the node models for three types of WPAN devices used for modeling 802.15.4scenarios. PAN Coordinator, GTS and Non GTS end device have the same node model asshown in Figure 2 (a) while the node model for analyzer is depicted in Figure 2 (b).

(a) (b)

Figure 2: Node Model (a) PAN Coordinator, GTS and Non GTS end device (b) Analyzer


20/30



As it has been observed from the Figure 2 (a), a node model for PAN Coordinator, GTS enddevice and Non GTS end device has three layers: physical, MAC and application layers. Physicallayer consists of a transmitter and a receiver compliant to the IEEE 802.15.4 specification,operating at 2.4 GHz frequency band and data rate equal to 250 kbps. MAC layer implementsslotted CSMA/CA and GTS mechanisms. The GTS data traffic coming from the application layeris stored in a buffer with a specified capacity and dispatched to the network when thecorresponding GTS is active. The non time-critical data frames are stored in an unbounded bufferand based on slotted CSMA/CA algorithm are transmitted to the network during the activeContention Access Period (CAP). This layer is also responsible for the generation of beaconframes and synchronizing the network when a given node acts as a PAN Coordinator. Finally isthe topmost application layer which is responsible for generation and reception of traffic consistsof two data traffic generators (i.e. Traffic Source and GTS Traffic Source) and one traffic sink. Thetraffic source generates acknowledged and unacknowledged data frames transmitted duringCAP. GTS traffic source can produce acknowledged and unacknowledged time-critical dataframes using GTS mechanism. The traffic sink module receives frames forwarded from lowerlayers. Figure 2 (b) shows the node model for the analyzer which consists of sink and a radioreceiver.Corresponding process models for PAN Coordinator, GTS end device, Non GTS end device andanalyzer that deals with each and every operation on the data are depicted in Figure 3:

(a)

(b)Figure 3: Process model (a) PAN Coordinator, GTS and Non GTS end device (b) Analyzer

Figure 3 (a) shows the process model for the PAN Coordinator, GTS and Non GTS end device. Itconsists of the various states: Init whose function is to initialize MAC and GTS scheduling;


21/30



Wait_beacon which is responsible for synchronizing the traffic of the node with rest of the WPANin order to minimize the collisions; Idle which is responsible for introducing delays in order tomake the maximum use of the resources; gts_slot which is responsible for generation, receptionand management of GTS traffic; Backoff_timer used for sensing the medium and transfer of data,CCA - for interrupt processing. Similarly figure 3 (b) shows the process model for analyzer whichconsists of init and idle states. Basically the process model explains how the data is sent from thegenerating node to the PAN Coordinator, taking into consideration the availability of PANCoordinator as it has to communicate with the other similar nodes.

Here three different Scenarios have been created with three different modulation formats like:BPSK, MSK and QAM_64. Following parameters have been set for these scenarios as shown inthe table 1 like: in GTS settings the value of GTS permit is common for all three types of devicesi.e. enabled.

Parameter \ Scenario PANCoordinator

GTS EnabledEnd Device

Non GTSEnd Device

Modulation BPSK, MSK, QAM_64Acknowledged Traffic Source

Destination MAC Address Broadcast PAN Coordinator

MSDU Interarrival Time

(sec)

Exponential(1.0) Constant (1.0) Exponential(1.0)

MSDU Size (bits) Exponential(912) Constant (0.0) Exponential(912)

Start Time (sec) 0.0 Infinity 1.0

Stop Time (sec) InfinityUnacknowledged Traffic Source

MSDU Interarrival Time(sec)

Exponential(1.0) Constant (1.0) Exponential(1.0)

MSDU Size (bits) Exponential(912) Constant (0.0) Exponential(912)

Start Time (sec) 0.1 Infinity 1.1

Stop Time (sec) InfinityCSMA/CA Parameters

Maximum Back-off Number 4

Minimum Back-offExponent

3

IEEE 802.15.4Device Mode PAN coordinator End Device

MAC Address Auto AssignedWPAN Settings

Beacon Order 14 7Superframe Order 6

PAN ID 0Logging

Enable Logging EnabledGTS Settings

GTS Permit Enabled

Start Time 0.0 0.1 Infinity

Stop Time Infinity

Length (slots) 1 0

Direction Receive TransmitBuffer Capacity (bits) 10,000 1000GTS Traffic Parameters

MSDU Interarrival Time(sec)

Exponential(1.0) Constant (1.0)

MSDU Size (bits) Exponential(912) Constant (0.0)

Acknowledgement Enabled DisabledTable 1: Parametric values for PAN Coordinator, GTS and Non GTS End Device in BPSK, MSK and

QAM_64 Scenarios


22/30



3. RESULTS AND DISCUSSIONS

Simulation has been carried out for the three different scenarios of IEEE 802.15.4 using QAM_64,MSK and BPSK. In this section results for the queue size at the radio transmitter have beenpresented and discussed for different types of devices in wireless sensor networks like: FullyFunctional Device (FFD) those devices that control the network and manage the routing tablesand communicate with each of the device in peer to peer mode, Reduced Functional Devices

(RFD) those devices which can only communicate to the FFD but not to each other.

Radio Transmitter Queue Size

3.1.1 FFD PAN CoordinatorFigure 4 below indicates the queue size at the radio transmitter of a PAN Coordinator. It isobserved that it is 0.2926, 0.2572 and 0.2261 packets for MSK, BPSK and QAM_64 respectively.It has been experimentally proved that queue size is maximum in case of MSK because itpurposefully generates the delays to reduce the phase shifts to produce amplifier-friendly signalswhich results in the long queues at the radio transmitter as compared to the other modulationschemes (e.g. BPSK, QAM_64 etc.) and also MSK has self synchronizing capability [17]. While ithas been observed that queue size is minimum in case of QAM_64 as it increases the efficiencyof transmission by utilizing both amplitude and phase variations [17, 23].

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0 132 264 396 528 660 792 924 1056 1188

Time (sec)

RadioT

ransm

itterQ

ueue

Size

(packets)

BPSK MSK QAM_64

Figure 4: Radio Transmitter Queue Size at PAN Coordinator

3.1.2 RFD GTS End DeviceFigure 5 indicates the queue size at the radio transmitter of a GTS end device. It is 0.0179,0.0020 and 0.0013 packets MSK, QAM_64 and BPSK respectively. It has been observed thatqueue size is maximum in case of MSK [17]. While it is minimum in case of BPSK as it canmodulate only 01 bit/sec and there is strong synchronization between the transmitter and the

receiver [23].


23/30



0

0.002

0.004

0.0060.008

0.01

0.012

0.014

0.016

0.018

0.02

0 132 264 396 528 660 792 924 1056 1188

Time (sec)

Radio

Transm

itterQueueSize

(pa

ckets)

BPSK MSK QAM_64

Figure 5: Radio Transmitter Queue Size at GTS End Device

3.1.3 RFD Non GTS End DeviceFigure 6 reveals the queue size at the radio transmitter of a Non GTS end device. It is 0.1621,0.1340 and 0.1172 packets for MSK, QAM_64 and BPSK respectively. It has been observed thatit is maximum in case of MSK [17], while it is minimum in case of BPSK [23].

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0 132 264 396 528 660 792 924 1056 1188

Time (sec)

Rad

io

Transm

itterQueueSize

(packets)

BPSK MSK QAM_64

Figure 6: Radio Transmitter Queue Size at Non GTS End Device

From the results obtained in figures: 4 for FFD and 5 & 6 for RFD (GTS & Non GTS), it has beenconcluded that if queue size at the PAN coordinator of 802.15.4 wireless sensor network is to betaken into consideration then QAM_64 should be preferred and if queue size at the GTS or NonGTS end device is to be considered then BPSK should be preferred.

4. CONSLUSION

This paper presents the queue size at the radio transmitter of 802.15.4 wireless sensor networkusing OPNET

Modeler 14.5. Here three different modulation scenarios for BPSK, MSK and

QAM_64 have been considered. Results reveals that queue size at the radio transmitter of PANCoordinator, GTS and Non GTS End Device is [0.2926, 0.2261, 0.2572], [0.0179, 0.0020, 0.0013]and [0.1621, 0.1340, 0.1172] packets for MSK, QAM_64 and BPSK respectively. It is concludedthat QAM_64 at the fully functional device and BPSK at the GTS and Non GTS RFDs should beimplemented if queue size at the radio transmitter of 802.15.4 WSN is to be minimized. Also it is


24/30



concluded that MSK at all type of devices in 802.15.4 for WSN is unsuitable as it results in thelarger queues as compared to the other modulation formats at all type of devices, as larger thequeues, larger will be the delays.

5. REFERENCES

[1] Jurcik, P., Koubaa, A., Alves, M., Tovar, E., Hanzalek, Z. A Simulation Model for theIEEE 802.15.4 protocol: Delay/Throughput Evaluation of the GTS Mechanism. Modeling,Analysis, and Simulation of Computer and Telecommunication Systems, 2007:MASCOTS '07. 15th International Symposium, 24-26 Oct. 2007.

[2] IEEE 802.15.4 OPNET Simulation Model: http://www.open-zb.net.[3] Patrick Hosein. Adaptive Algorithm for Mapping Channel Quality Information to

Modulation and Coding Schemes. IEEE: 2009.[4] Yafei Hou, Masanori Hamamura, Shiyong Zhang. Performance Tradeoff with Adaptive

Frame Length and Modulation in wireless Network. Proceedings of the IEEE 2005, theFifth International Conference on Computer and Information Technology (CIT 05).

[5] R. Schober, W. H. Gerstacker, L. Lampe. On suboptimum receivers for DS-CDMA withBPSK modulation. Signal Processing 85 (2005): 1149 1163, Elsevier.

[6] Rajarshi Mahapatra, Anindya Sunder Char, Debasish Datta. Dynamic Capacity

Allocation for Voice and Data Using Adaptive Modulation in Wireless Networks. IEEE:2006.

[7] Jan Magne Tjensvold. Comparision of the IEEE 802.11, 802.15.1, 802.15.4 and802.15.6 wireless standards. IEEE: September 18, 2007.

[8] Jason Lowe, Advanced Upstream Modulation. Clearcable Technical Summit: July 27,2007.

[9] Feng Chen, Nan Wang, Reinhard German, Falko Dressler. Simulation study of IEEE802.15.4 LR-WPAN for industrial applications. Wireless Communications and MobileComputing: 2009.

[10] S.M. Mahdi Alavi, Martin J. Hayes. Robust Active Queue management design: A loop-shaping approach. Elsevier: Computer Communications, 2009.

[11] Shan Chen, Brahim Bensaaou. Can high-speed networks survive with DropTail queuemanagement. Elsevier: Computer Networks, 2006.

[12] P. Venkata Krishna, Sudip Misra, Mohammad S. Obaidat, V. Saritha. An efficientapproach for distributed dynamic channel allocation with queues for real-time and non-real-time traffic in cellular networks. Elsevier: The Journal of Systems and Software,2009.

[13] Subhash Nanjunde Gowda. Minimum shift keying. Spread Spectrum Systems: 24 May2004.

[14] Qiuyan Xia, Xing Jin, Mounir Hamdi. AQM with Dual Virtual PI Queues for TCPUplink/Downlink Fairness in Infrastructure WLANs. IEEE: 2007.

[15] Jun Luo, Lingge Jiang, Chen He, Finite Queuing Model Analysis for Energy and QoSTradeoff in Contention-Based Wireless Sensor Networks, IEEE, 2007.

[16] Yi Fan, Zhong-Ping Jiang, Hao Zhang. Stablizing Queues in Large-scale Networks.IEEE: 2005.

[17] Charan Langton. Intuitive Guide to principles of Communications.

www.complextoreal.com Dec 2005.[18] Prasan Kumar Sahoo, Jang-Ping Sheu. Modelling IEEE 802.15.4 based WirelessSensor Network with Packet Retry Limits. ACM: PE-WASUN08.

[19] Jelena Misic, Vojislav B. Misic. Queuing Analysis of Sleep Management in an 802.15.4Beacon Enabled PAN. IEEE.

[20] L. Chisci, R. Fantacci, L. Mucchi, T. Pecorella. A Queue-Based Approach to PowerControl in Wireless Communication Networks. IEEE: 2008.

[21] Shahram Teymori, Weihua Zhuang. Finite Buffer Queue Analysis and Scheduling forHeavy-tailed Traffic in Packet-Switching wireless Networks. IEEE: 2005.


25/30



[22] Omesh Tickoo, Biplab Sikdar. Queuing Analysis and Delay Mitigation in IEEE 802.11Random access MAC based Wireless Networks. IEEE: 2004.

[23] http://www.en.wikipedia.org.


26/30

CALL FOR PAPERS

Journal: International Journal of Computer Networks (IJCN)Volume: 2 Issue: 5ISSN:1985-4129

URL: http://www.cscjournals.org/csc/description.php?JCode=IJCN

About IJCN

The International Journal of Computer Networks (IJCN) is an archival,bimonthly journal committed to the timely publications of peer-reviewed andoriginal papers that advance the state-of-the-art and practical applications ofcomputer networks. It provides a publication vehicle for complete coverageof all topics of interest to network professionals and brings to its readers thelatest and most important findings in computer networks.

To build its International reputation, we are disseminating the publicationinformation through Google Books, Google Scholar, Directory of Open AccessJournals (DOAJ), Open J Gate, ScientificCommons, Docstoc and many more.Our International Editors are working on establishing ISI listing and a goodimpact factor for IJCN.

IJCN List of Topics

The realm of International Journal of Computer Networks (IJCN) extends, butnot limited, to the following:

Algorithms, Systems andApplications

Ad-hoc Wireless Networks

ATM Networks Body Sensor Networks Cellular Networks Cognitive Radio Networks

Congestion and Flow Control Cooperative Networks Delay Tolerant Networks Fault Tolerant Networks

Information Theory Local Area Networks Metropolitan Area Networks MIMO Networks Mobile Computing Mobile Satellite Networks

Multicast and Broadcast Networks Multimedia Networks Network Architectures and Protocols Network Coding

Network Modeling and Performance

Analysis Network

Network Operation and

Management Network Security and Privacy Network Services and Applications

Optical Networks Peer-to-Peer Networks Personal Area Networks Switching and Routing Telecommunication Networks Trust Worth Computing Ubiquitous Computing Web-based Services Wide Area Networks Wireless Local Area Networks

Wireless Mesh Networks Wireless Sensor Networks


27/30

Important Dates

Volume: 2

Issue: 5

Paper Submission: September 30 2010

Author Notification: November 01, 2010Issue Publication: November / December 2010


28/30

CALL FOR EDITORS/REVIEWERS

CSC Journals is in process of appointing Editorial Board Members forInternational Journal of Computer Network (IJCN). CSC Journalswould like to invite interested candidates to join IJCN network ofprofessionals/researchers for the positions of Editor-in-Chief, Associate

Editor-in-Chief, Editorial Board Members and Reviewers.

The invitation encourages interested professionals to contribute intoCSC research network by joining as a part of editorial board members

and reviewers for scientific peer-reviewed journals. All journals use anonline, electronic submission process. The Editor is responsible for the

timely and substantive output of the journal, including the solicitationof manuscripts, supervision of the peer review process and the final

selection of articles for publication. Responsibilities also includeimplementing the journals editorial policies, maintaining high

professional standards for published content, ensuring the integrity ofthe journal, guiding manuscripts through the review process,

overseeing revisions, and planning special issues along with theeditorial team.

A complete list of journals can be found at

http://www.cscjournals.org/csc/byjournal.php. Interested candidatesmay apply for the following positions through

http://www.cscjournals.org/csc/login.php.

Please remember that it is through the effort of volunteers such as

yourself that CSC Journals continues to grow and flourish. Your helpwith reviewing the issues written by prospective authors would be very

much appreciated.

Feel free to contact us at [email protected] if you have anyqueries.


29/30

Contact Information

Computer Science Journals Sdn BhDM-3-19, Plaza Damas Sri Hartamas

50480, Kuala Lumpur MALAYSIA

Phone: +603 6207 1607

+603 2782 6991Fax: +603 6207 1697

BRANCH OFFICE 1

Suite 5.04 Level 5, 365 Little Collins Street,MELBOURNE 3000, Victoria, AUSTRALIA

Fax: +613 8677 1132

BRANCH OFFICE 2

Office no. 8, Saad Arcad, DHA Main BulevardLahore, PAKISTAN

EMAIL SUPPORTHead CSC Press: [email protected]

CSC Press: [email protected]: [email protected]


30/30

International Journal of Computer Networks (IJCN) Volume 2, Issue 4, 2010

Documents

Transcript of International Journal of Computer Networks (IJCN) Volume 2, Issue 4, 2010